Arduino Nano works only one time and then the computer does not recognize it

I need another solution. The device will be in an enclosure with no access to the reset button.

That suggests some sort of hardware damage then, during that initial connect/disconnect

It may be a design fault or bad component choice that you can't do anything about. The only thing I can think of that you may be able to influence, is the order in which the USB pins make contact. Which should be defined by the USB standard, but if hypothetically you had an out-of-spec Chinese copy of a USB port it may be connecting/disconnecting in the wrong order.

If there was something like that going on, you might be able to avoid the problem (with new Nanos that have not already been 'blown') by disconnecting from the PC end, where you are presumably using a known good host port and cable. No reason at all to think this is at all likely, but it's the only thing I can think of that you have any chance to influence, so you have nothing to lose by trying.

szangvil:
I need another solution. The device will be in an enclosure with no access to the reset button.

If you really had to you could run a switch to the reset and gnd pins and place the switch so its accessible from outside the enclosure.

Is it also a chinese ftdi? The reason i ask is because it sounds to me like a driver problem. Ive had tons of weird usb related problems like this in the past. Sometimes the pc wouldnt recognize the device even after rebooting a dozen times and then suddenly everything would work.

Hello,
I bought a Nano from Littlebird electronics and another through ebay. The first one I have uploaded to it many of times and had some trouble with data corruption so I bought the other one to ensure that it was not the chips fault. To my discovery it was my software that was causing the issue. Any way my point is that I have had 1 nano which I have uploaded maybe 50 times and another also 50 times now as I keep swapping and changing. Now I can not connect to either of them because my windows 7 keeps telling me that the device has not installed correctly. I have pointed windows to the Arduino drivers and still no solution. This has been happening for a while on some of my usb ports but if I changed to a different port it would install correctly. Now all of my ports (4 of them) will not install the correct drivers. What is going on here it seems like a common problem but do not see any solutions. Why has this suddenly happened, is it windows update that has caused this?

Have you tried deleting the old/failed/brokendevices via Hardware manager, and uninstalling the driver? It would prompt to reinstall via PnP when you reconnect the Arduino.

Ok, here is the thing:

I have 20 of them. Some are labeled "Funduino" and some "Arduino".
After I reboot my computer, all of the sudden ALL of them work the first time I connect them. If I disconnect and re-connect, about 10 will not be recognized by windows.
The thing is, if one is not recognized by my computer and I connect it to another, it still will not recognized...

Anyway, I am baffled...

I was pointed by Robert that the FT232R requires the TEST pin to be connected to GND. I tried to do that on one of the Arduinos that was "acting up" (i.e., first connection is successful but successive connections fail). That seems to make a successful connection every time.
This is not a workable solution for me since they are all soldered to my PCB and I don't have a the tools to remove them.

If this is the solution, why are the Arduino Nano schematics and design not updated?

szangvil:
Ok, here is the thing:

I have 20 of them. Some are labeled "Funduino" and some "Arduino".
After I reboot my computer, all of the sudden ALL of them work the first time I connect them. If I disconnect and re-connect, about 10 will not be recognized by windows.
The thing is, if one is not recognized by my computer and I connect it to another, it still will not recognized...

Anyway, I am baffled...

I was pointed by Robert that the Atmega328 requires the TEST pin to be connected to GND. I tried to do that on one of the Arduinos that was "acting up" (i.e., first connection is successful but successive connections fail). That seems to make a successful connection every time.
This is not a workable solution for me since they are all soldered to my PCB and I don't have a the tools to remove them.

If this is the solution, why are the Arduino Nano schematics and design not updated?

Where is this 'TEST' pin you speak of? I don't see it on the nano version 3.0 schematic drawing? Got a link?

Lefty

Im sorry, my mistake. It's on the FT232R chip. Pin 26.
I'v edited my previous post with this change.

szangvil:
Im sorry, my mistake. It's on the FT232R chip. Pin 26.
I'v edited my previous post with this change.

That surely must be a mistake on the Nano's V3 schematic as the classic arduino Duemilanove uses the FTDI chip and shows pin 26 being grounded.

And from the FTDI datasheet, it says this about pin 26:

26 TEST Input
Puts the device into IC test mode. Must be tied to GND for normal
operation, otherwise the device will appear to fail.

Lefty

I did see this solution (pin grounding) on another post Arduino Forum but in my case I have I programmed my nano repeatedly for months and now all of a sudden I am getting this error on 2 different Nanos, using multiple different cables. If it was something wrong with the nano circuitry then why has it worked for so long?
I can get it working again by uninstalling the driver and plugging the device back in which I had to do on each usb port, but sometimes it will still not recognize and I have to go through the uninstall process again. Will keep you posted if I find anything.

Typhoon, I had about 80 Nanos over the past few months. In last batch of 20, some "act funny" and seem to be recognized by Windows randomly while others are fine.
As I said before, I "sacrificed" on Nano and did the TEST-GND blob bridge and that fixed the random behavior - windows recognized it all the time after plugging in and out tens of times (repeatedly).

Update:

Plugin the first time works on all 20pcs. If I plug in and out very fast (a few seconds in between), about 10 will not be recognized.
If I plug in and out with a delay of over 2 minutes, they all work... beats me.

Ok you convinced me :slight_smile: I will wire Test to GND, its going to be like performing heart surgery on those SMD pins.

I was "afraid" to do it as well, but solder flows to the right place all on its own, so don't worry :slight_smile:

Another happy user which has fixed his Nano just a while ago. It should be written about the problem in capital letters on the official Arduino Nano page, and the schematics/boards updated.

I know this topic is really old, but I was having the same issue today with chinese nanos, and just got it solved by installing the original FTDI drivers manually like in this video, now it recognize the nanos perfectly every time i plug it.

That's the video I used to unbrick all of mine. Takes about 3 minutes of driver manipulation.

i have had a similar problem recently with windows 7.
from what i am able to find out, it is a fake FTDI chip which is causing all the problems, so windows doesn't recognise it as an FTDI device. I will have a hunt around, there was a fix, i think it was a different driver. I also have a linux machine which works perfectly, looks like good old windows trying to tell you what you have and getting it wrong again!

adamants:
Looks like good old windows trying to tell you what you have and getting it wrong again!

No, it's kind of worse than that!
(at 1:20 :grinning: )
A driver from FTDI - published in the Windows Updates and subsequently removed as being of criminal intent, deliberately disables the counterfeit chips, however it is fixable.

The take-home messages are:

  • Don't use Windoze unless you are forced to do so by badly written (and it generally is) proprietary software that will not work under Linux. Linux uses generic drivers that simply - work!
    Reliably

  • Do not use FTDI chips in project designs.

That's a pretty broad generalization Paul__B.
Mixing Windows drivers on Linux OS is likely a problem.

There is no problem using actual FTDI chips purchased from legit sources in projects. I have been doing so since 2010 with no issues, and will continue to do so. Questionable (i.e. fake, counterfeit) chips from questionable sources (low-ball priced parts from ebay) may present problems. As an electrical engineer who designed boards in industry for many years, and having seen many a report of counterfeit parts causing issues and requiring expensive post-fielding replacement/repairs, I will not purchase ANY ICs from e-bay suppliers. These fake chips may not work outright, may be programmed incorrectly (the real parts are programmable), or may not work with (now) older FTDI drivers. The driver from FTDI was their attempt to protect their intellectual property, I don't have a problem with that.
"Criminal intent", that could be grounds for libel and/or slander.